Build with AI without leaving the thinking to it.
Engineering With AI is the method we developed while building Backstory. It’s for the work beyond a promising demo: understanding the problem, shaping a change, checking what was built and leaving the next session enough context to continue.
Engineering With AI is available now. Thinking With AI and Leading IT With AI are coming later in 2026.
Read a little before you decide.
From Chapter 1, From Idea to Intent.
The Ten People in a Room
Here’s a thought experiment that reveals the problem.
Imagine you gather ten people in a room and describe your idea for an application. You give them the same verbal description, or maybe two or three sentences about what you want to build. Then you hand each person a whiteboard marker and ask them to sketch what they just heard.
You’ll get ten different designs.
Not variations on a theme. Fundamentally different interpretations. One person draws a mobile app. Another draws a desktop dashboard. Someone focuses on the admin panel. Someone else draws the customer-facing portal. The person in the corner has added three features you never mentioned because “obviously you’d need those.”
This is the alignment problem, and it exists inside your own head too.
Later in the same chapter:
AI as a Thinking Brake
The first real value of AI isn’t that it helps you move faster. It’s that it can help you slow down without losing momentum.
Before you write a single line of code, use the AI to interrogate your idea. You want to move from a vague “Idea” to a sharp “Intent.”
Intent isn’t a feeling. It’s a clear, actionable directive that defines the boundaries of your work. If you can’t articulate your intent, you’re never going to be able to explain it to the AI, and the AI will fill the vacuum with its own assumptions. You likely won’t notice until you’re deep in code that’s actually solving the wrong problem.
Then put it to work in your own project.
The book follows the engineering work, not just the conversation. You’ll develop the decisions and records that make a project easier to build, review and return to.
The harness is the companion system. The book explains the reasoning behind the method; the harness gives that work a home in your repository.
- Decide what you actually need.Develop intent, challenge assumptions and describe the people and outcomes the work is for.
- Make the change reviewable.Break work into Feature Capsules, use SPECS to keep context and apply human Ready and Done decisions.
- Check what happened.Use review, tests and evidence to examine the result, then retain the learning for the next change.
Building is only part of working with AI.
Two companion books are coming soon. They’re not available to order yet.
Thinking With AI
For the question that isn’t clear yet, the assumption that needs testing and the decision that deserves more than the first polished answer. Explore Partnered Intelligence, reasoning modes and the human judgement around the conversation.
Later in 2026.
Leading IT With AI
For leaders coordinating people and AI across strategy, services, suppliers, risk and change. Develop how you delegate, review and make decisions without losing responsibility for the people affected.
Later in 2026.
Get Engineering With AI.
Digital edition · £39. No VAT added.
Ready to use the method in your own project? Buy the digital edition through secure checkout. The PDF and EPUB are included, and your order and download links will be available in My account.
Thinking With AI and Leading IT With AI are coming soon and aren’t available to buy yet.